I should repost the reply I sent on the c4lcon list here:

Hey Ed,

Thanks for posting this summary here. It's really cool to see a
description of how this is working.

I think this is a pretty good example of how a library data mgt
interface of the future might work:
* Grab some free text describing a thing;
* Try to clean it up, extract important concepts / themes topics
* Reconcile against some sort lod-lam friendly controlled vocabulary/ies
* Offer "cataloger" types an interface to accept / reject / refine
those mappings as well as the text of the "metadata" itself.

I would go to a breakout session about this.


On Wed, Feb 1, 2012 at 9:46 AM, Cynthia Ng <[log in to unmask]> wrote:
> Just a quick 2 cents. I only found out about the feed by reading this
> conversation. I think it would be great to make the RSS link a little
> more obvious from the front page.
> -Cynthia
> On Wed, Feb 1, 2012 at 8:22 AM, Michael J. Giarlo
> <[log in to unmask]> wrote:
>> I smell a potential breakout session.
>> -Mike
>> P.S. No, really, jokers, that's what I smell.
>> On Jan 31, 2012 11:30 PM, "Ed Summers" <[log in to unmask]> wrote:
>>> I guess it's rarely a good idea to respond to your own post, but I
>>> forgot to add that when a job is "published" on it
>>> will show up in the site's Atom feed [1]. The feed should be usable by
>>> your feed reader of choice, and could also be useful if you want to
>>> syndicate the jobs elsewhere.
>>> //Ed
>>> [1]
>>> PS. It was kind of fun to finally use the "tag" link relation to mark
>>> up the job tags in the feed with Freebase URLs. For example:
>>> <entry>
>>>        ...
>>>        <link rel="tag" title="Unix"
>>> href="" type="text/html" />
>>>        <link rel="tag" title="Unix [JSON]"
>>> href=""
>>> type="application/json" />
>>>        <link rel="tag" title="Unix [RDF]"
>>> href="" type="application/rdf+xml"
>>> />
>>> </entry>

Corey A Harper
Metadata Services Librarian
New York University Libraries
20 Cooper Square, 3rd Floor
New York, NY 10003-7112
[log in to unmask]